During my tenure at Cloudmendx as a medical service provider, focusing on developing in-depth AI solutions for insurance cost and resources forecasting and medicare management platform. We had successful project partner with Kaiser, Sharecare, Epic, Lyrics.
1.Refactoried the php based service by reorganize frontend with react, backend with. Laravel, organize api with graphql and redesigned data flow to reduce 73% response time to improve user experience.
2.Design ETL flow and infrastructure with client to migrating external data into database, and designed schema with proper normalizing, shading and indexing method to minimizing processing time.
3.Refactoring docker images to reduce vulnerability check and dependencies compatibility from scratch. designed and lead the integration of Kubernetes pods to develop a IaaS products for scalability, cost-efficiency and liability(generally 36% cost lower than hosting with original infrastructure).
4.Designed a Supervised Learning workflow, involving data preprocessing to format and label data appropriately. proficiently utilize Conda for environment configuration and GPU driver management. Additionally, I employ Scikit-Learn and PyTorch for ML, concluding with validation using k-score and reach 95% precision on ICD10 code recognition from medical notes.
5.Designed modalized products with multiple microservices, leaded TDD(test driven design) workflow, deployed and monitored service on AWS with high avability.
6.Established and managed documents convention(URD, Issue tracking, feasibility report, API documents, Test report and UML diagrams) with JIRA confluence, Postman, draw.io and markdown.
7.Setup Zabbix sever and proxy to help manage Azure resource and collecting data for services management
